Boox V.] EMBRYOLOGY AND PALHIONTOLOGY. — 625 
order of succession of plants, and the absence of transitional forms among 
them. Ferns, equisetums, and lycopods appear as far back as the Old 
Red Sandstone, not in simple or more generalized, but in more complex 
structures than their living representatives. ‘T'he earliest known conifers 
were well-developed trees with woody structure and fruits as highly 
differentiated as those of their living representatives. The oldest 
dicotyledons yet found, those of the upper Cretaceous formations, contain 
representatives of the three great divisions of Apetale, Monopetale, and 
Polypetalz, in the same deposit. These ‘are not generalized types, but 
differentiated forms which, during the intervening epochs, have not 
developed even into higher generic groups.” * 
Professor A. Agassiz has recently drawn attention to the parallelism 
between embryonic development and paleontological history. Taking 
the sea-urchins as an illustrative group, he points out the interesting 
analogies between the immature conditions of living forms and the 
appearance of corresponding phases in fossil genera. He admits, how- 
ever, that no early type has yet been discovered whence star-fishes, 
sea-urchins, or ophiurans might have sprung; that the several orders of 
echinoderms appear at the same time in the geological record, and that 
it is impossible to trace anything like a sequence of genera or direct 
filiation in the paleontological succession of the echinids, though he 
does not at all dispute the validity of the theory which regards the 
present echinids as having come down in direct succession from those of 
older geological times. In the case of the numerous genera which have 
continued to exist without interruption from early geological periods 
and have been termed “persistent types,” it is impossible not to admit 
that the existing forms are the direct descendants of those of former ages. 
If, then, some genera have unquestionably been continuous, the evolu- 
tionist argues, it may reasonably be inferred that continuity has. been 
the law, and that even where the successive steps of the change cannot 
be traced, every genus of the living world is genetically related to other 
genera now extinct. 
Among the fossil mammalia many indications have been pointed out 
of an evolution of structure. Of these, one of the best known and most 
striking is the genealogy of the horse, as worked out by Professor 
O. C. Marsh.2 The original, and as yet undiscovered, ancestor of our 
modern horse had five toes on each foot. In the oldest known equine 
type (EHohippus—an animal about the size of a fox, belonging to 
the early part of the Eocene period) there were four well-developed 
toes, with the rudiment of a fifth, on each fore-foot, and three on each 
hind foot. Ina later part of the same geological period appeared the 
Orohippus, a creature of about the same size, but with only four toes in 
front and three behind. ‘Traced upwards into younger divisions of the 
Tertiary series, the size of the animal increases, but the number of 
digits diminishes, until we reach the modern Hquus, with its single toe 
and rudimentary splint-bones. 
Another remarkable example, that of the camels, is cited by 
Professor H. D. Cope. The succession of genera is seen in the same 
1 Carruthers, Geol. Mag. 1876, p. 362. 
2 Ann. Mag. Nat. Hist. Nov. 1880, p. 369. “Report on Echinoidea,” Challenger 
Expedition, vol. iii. p. 19. 
3 Amer. Journ. Sci. 1879, p. £99. 
28 
